Handling and analyzing to web financial data is becoming a challenge issue in knowledge management and education to accounting practitioners. eXtensible Business Reporting Language (XBRL), which is a type of web financial reporting, describes and recognizes financial items by tagging metadata. The goal is to make it possible for financial reports to be handled accurately and to be analyzed automatically via human-computer interactive approach. To solve the issues that analyzing financial data might create, (1) a decidable Tableau algorithm for description logic DLRBR is proposed; (2) a formalization approach about part-whole relationship in XBRL is proposed; (3) consistency checking and logical reasoning of XBRL metadata is discussed; (4) then a human-computer interactive prototype system is developed to verify the proposed method. The results can help prompting the ability of knowledge management and education in accountings.